Legal claims defining the scope of protection. Each claim is shown in both the original legal language and a plain English translation.
1. A computer-implemented method for training and using a neural network for subcluster classification, the method comprising: receiving or generating, by one or more processors, a plurality of user data sets of users, wherein each user data set in the plurality of user data sets comprises a user identification data of a user and a detailed user data of the user; grouping the plurality of the user data sets, by the one or more processors, into one or more clusters of user data sets; grouping, by the one or more processors, each of the one or more clusters into a plurality of subclusters; for each of the plurality of subclusters, training the neural network, by the one or more processors, to associate the subcluster with one or more sequential patterns found within the subcluster, based on one or more user data sets in the subcluster of the plurality of user data sets, to generate a trained neural network; receiving a first series of transactions of a first user; inputting, by the one or more processors, the first series of transactions into the trained neural network; classifying, using the trained neural network, by the one or more processors, the first user into a subcluster of the plurality of subclusters to identify a classified subcluster of the first user, based on the first series of transactions of the first user input into the trained neural network; based on the classified subcluster of the first user, determining, by the one or more processors, whether a model user associated with the first user should be identified; and based on the determination of whether the model user associated with the first user should be identified, searching, by the one or more processors, the plurality of user data sets for the user data set of the model user, wherein the user data set of the model user (i) is grouped into one of one or more model subclusters, wherein the one or more model subclusters is predetermined to be model subclusters among the plurality of subclusters, and (ii) comprises a second series of transactions having one or more transactions with identical attributes, compared to attributes of one or more transactions of the first series of transactions.
2. The method of claim 1 , wherein the grouping of the plurality of the user data sets into one or more clusters of user data sets is based on one or more of annual income, education level, family size, or job category of the users.
This invention relates to a method for grouping user data sets into clusters based on demographic and socioeconomic factors to improve data analysis and decision-making. The method involves collecting user data sets, each containing information about individual users, and grouping these data sets into clusters based on shared characteristics. The grouping is performed using one or more criteria, including annual income, education level, family size, or job category of the users. By analyzing these factors, the method enables more accurate segmentation of users, allowing for targeted insights, personalized recommendations, or tailored marketing strategies. The clustering process helps identify patterns and correlations within the data, which can be used for various applications, such as financial services, education, workforce planning, or consumer behavior analysis. The method ensures that the grouping is based on relevant and meaningful attributes, enhancing the reliability and usefulness of the clustered data for further analysis. This approach improves decision-making by providing a structured way to categorize users and understand their needs or behaviors more effectively.
3. The method of claim 2 , wherein the grouping of the each of the one or more clusters into a plurality of subclusters is based on one or more of credit score, an account balance, available credit, or percentage of credit used of users in each of the one or more clusters.
This invention relates to a method for analyzing and grouping user data, particularly in financial or credit-related applications. The method addresses the challenge of efficiently categorizing users into meaningful segments for targeted analysis or decision-making, such as credit risk assessment or personalized financial services. The method involves clustering users into groups based on shared characteristics, then further subdividing those groups into subclusters. The subdivision is based on specific financial metrics, including credit score, account balance, available credit, or the percentage of credit used by users within each cluster. By refining the groupings using these metrics, the method enables more precise segmentation, allowing for better risk assessment, personalized offers, or tailored financial recommendations. The initial clustering step organizes users into broad categories, while the subclustering step applies additional filters to refine those categories. This two-tiered approach ensures that users are grouped not just by general similarities but also by specific financial behaviors or statuses, leading to more accurate and actionable insights. The method is particularly useful in financial institutions, credit bureaus, or any system requiring detailed user segmentation for decision-making.
4. The method of claim 1 , wherein the step of receiving or generating a plurality of user data sets of users further comprises: removing, by the one or more processors, personally identifiable information from each of the plurality of user data sets.
This invention relates to data processing systems that handle user data sets, particularly focusing on privacy protection by removing personally identifiable information (PII). The problem addressed is the need to process user data while ensuring compliance with privacy regulations and protecting sensitive information. The method involves receiving or generating multiple user data sets, each containing information about individual users. A key step in this process is the removal of personally identifiable information from each data set. This removal is performed by one or more processors to ensure that the data cannot be traced back to specific individuals. The processors may use techniques such as anonymization, pseudonymization, or redaction to strip out identifiers like names, addresses, or unique identifiers. The processed data sets are then used for further analysis or applications while maintaining user privacy. This approach is particularly useful in scenarios where data must be shared or analyzed without exposing personal details, such as in healthcare, finance, or market research. By systematically removing PII, the method ensures that the data remains useful for analysis while mitigating privacy risks. The invention may be integrated into larger data processing systems or privacy-focused software solutions.
5. The method of claim 1 , wherein the method further comprises: based on both the first series of transactions and the second series of transactions, generating, by the one or more processors, an indicator of a resolution associated with the first series of transactions.
This invention relates to transaction processing systems, specifically methods for analyzing and resolving discrepancies between transaction records. The problem addressed is the difficulty in accurately identifying and resolving inconsistencies that arise when comparing two sets of transaction data, such as those from different sources or systems. Such discrepancies can lead to financial errors, compliance issues, or operational inefficiencies. The method involves processing a first series of transactions and a second series of transactions, where these series may originate from different systems or time periods. The system compares the two series to detect mismatches or inconsistencies. Based on this comparison, the system generates an indicator of a resolution for the first series of transactions. This resolution indicator may include a flag, a status, or a recommendation for corrective action, helping users or automated systems address the discrepancies efficiently. The resolution process may involve reconciliation, correction, or flagging transactions for further review. The method ensures that transaction data integrity is maintained by providing clear, actionable insights into discrepancies, reducing manual effort and improving accuracy in financial and operational reporting.
6. The method of claim 5 , wherein the generating the indicator of the resolution further comprises converting the indicator into a natural language statement.
This invention relates to a method for generating an indicator of resolution in a technical or problem-solving context, particularly for converting technical indicators into natural language statements. The method addresses the challenge of making technical data or outcomes more accessible and interpretable by non-experts or users who may not be familiar with specialized terminology or formats. The method involves generating an indicator of resolution, which represents the outcome or solution to a technical problem or process. This indicator is then converted into a natural language statement, making it easier to understand and communicate. The conversion process ensures that the technical details are accurately translated into plain language, preserving the meaning while improving clarity. The method may be applied in various domains, such as software development, engineering, or data analysis, where technical results need to be conveyed to stakeholders who lack specialized knowledge. By converting technical indicators into natural language, the method enhances collaboration, decision-making, and reporting processes. The invention may also include additional steps, such as analyzing the technical data to determine the appropriate resolution indicator, validating the conversion process to ensure accuracy, and customizing the natural language output based on the target audience or context. The method ensures that technical information remains precise while being presented in a user-friendly format.
7. The method of claim 6 , wherein the method further comprises: displaying, by the one or more processors, the natural language statement on a device associated with the first user.
This invention relates to a method for processing and displaying natural language statements in a communication system. The method addresses the challenge of efficiently conveying information between users in a way that is both accurate and easily accessible. The system involves receiving a natural language statement from a first user, where the statement is generated based on input data such as text, speech, or other forms of communication. The method then processes this statement to ensure clarity and relevance before transmitting it to a second user. Additionally, the method includes displaying the processed natural language statement on a device associated with the first user, allowing the sender to review the content before transmission. This ensures that the information is correctly formatted and understood before being shared with the intended recipient. The method may also involve verifying the statement for accuracy, context, or other criteria before display or transmission. The system is designed to improve communication efficiency and reduce errors in information exchange.
8. A computer system for training and using a recurrent neural network for subcluster classification, the computer system comprising: a memory having processor-readable instructions stored therein; and at least one processor configured to access the memory and execute the processor-readable instructions, which when executed by the processor configures the processor to perform a plurality of functions, including functions for: receiving or generating a plurality of user data sets, wherein each user data set of the plurality of user data sets comprises a user identification data of a user and a detailed user data of the user; grouping the plurality of the user data sets into one or more clusters of user data sets; grouping each of the one or more clusters into a plurality of subclusters; for each of the plurality of subclusters, training the recurrent neural network to associate the subcluster with one or more sequential patterns found within the subcluster, based on the user data sets in the subcluster, to generate a trained recurrent neural network; receiving a first series of transactions of a first user; inputting the first series of transactions into the trained recurrent neural network; classifying, using the trained recurrent neural network, the first user into a subcluster of the plurality of subclusters, based on the first series of transactions input into the trained recurrent neural network, to qenerate a classified subcluster of the first user; based on the classified subcluster of the first user, determining whether a model user associated with the first user should be identified; and if the determination of whether the model user associated with the first user should be identified is that the model user should be identified, searching the plurality of user data sets for the user data set corresponding to the model user, wherein the user data set of the model user (i) is grouped into one of one or more model subclusters, wherein the one or more model subclusters is predetermined to be model subclusters among the plurality of subclusters, and (ii) comprises a second series of transactions having one or more transactions with identical attributes, compared to attributes of one or more transactions of the first series of transactions.
This invention relates to a computer system for training and using a recurrent neural network (RNN) to classify users into subclusters based on transaction patterns. The system addresses the challenge of identifying model users whose transaction behavior closely matches that of a target user, enabling targeted recommendations or interventions. The system processes user data sets, each containing user identification and detailed transaction information. These data sets are grouped into clusters and further divided into subclusters. For each subcluster, the RNN is trained to recognize sequential transaction patterns specific to that subcluster. When a new user's transaction series is input, the RNN classifies the user into the most relevant subcluster. If the classified subcluster is a predetermined model subcluster, the system searches for a model user whose transaction attributes match those of the target user. This allows for the identification of users with similar behavior, facilitating personalized actions such as recommendations or fraud detection. The system leverages RNNs to capture temporal dependencies in transaction sequences, improving the accuracy of user classification and model user identification.
9. The system of claim 8 , wherein the plurality of functions further comprise: if the determination of whether the model user associated with the first user should be identified is that the model user should be identified, generating an indicator of a resolution associated with the first series of transactions, based on both the first series of transactions and the second series of transactions.
This invention relates to a system for analyzing user transactions to identify model users and generate resolution indicators. The system addresses the challenge of determining whether a user's transaction patterns align with those of a model user, and if so, generating a resolution indicator based on both the user's transactions and the model user's transactions. The system includes a processor and a memory storing instructions that, when executed, cause the processor to perform a series of functions. These functions include receiving a first series of transactions associated with a first user and a second series of transactions associated with a model user. The system then determines whether the model user should be identified for the first user based on the first and second series of transactions. If the determination is that the model user should be identified, the system generates an indicator of a resolution associated with the first series of transactions. This resolution indicator is based on both the first series of transactions and the second series of transactions, allowing for a more accurate and informed resolution. The system may also include additional functions, such as receiving a third series of transactions associated with a second user and determining whether the model user should be identified for the second user based on the third series of transactions. If the model user is identified for the second user, the system generates an indicator of a resolution associated with the third series of transactions, again based on both the third series of transactions and the second series of transactions. This allows the system to analyze multiple users' transactions in relation to a model user, providing a scalable solution for transaction analysis and res
10. The system of claim 9 , wherein the generating the indicator of the resolution further comprises converting the indicator into a natural language statement.
A system for generating an indicator of resolution in a technical or problem-solving context converts the indicator into a natural language statement. The system processes input data related to a technical problem or task, analyzes the data to determine a resolution or solution, and generates an indicator representing the resolution. This indicator is then transformed into a human-readable natural language statement, allowing users to easily understand the resolution without requiring technical interpretation. The system may involve machine learning, natural language processing, or other computational techniques to analyze the input data and generate the resolution indicator. The conversion to natural language ensures clarity and accessibility, making the system suitable for applications in technical support, automated troubleshooting, or decision-making systems where human-readable output is essential. The system may integrate with existing databases, user interfaces, or other systems to provide seamless resolution communication. The natural language statement may include details such as the nature of the problem, the proposed solution, and any steps required to implement the resolution. This approach enhances usability and reduces the need for specialized knowledge to interpret the resolution.
11. The system of claim 10 , wherein the plurality of functions further comprises: if the determination of whether the model user associated with the first user should be identified is that the model user should be identified, displaying the natural language statement on a device associated with the first user.
This invention relates to a system for identifying and displaying natural language statements to users based on model user associations. The system operates in a domain where users interact with digital platforms, and the problem addressed is determining whether a model user should be identified and, if so, presenting relevant information to the first user. The system includes a plurality of functions that analyze user data to make this determination. If the analysis concludes that the model user should be identified, the system displays a natural language statement on a device associated with the first user. The natural language statement is generated based on the model user's attributes or actions, providing contextually relevant information to the first user. The system may also include functions for processing user input, comparing user data to model user profiles, and dynamically updating the displayed information. The overall goal is to enhance user experience by delivering personalized or contextually appropriate messages based on predefined model user criteria.
12. The system of claim 8 , wherein the grouping of the plurality of the user data sets into one or more clusters of user data sets is based on one or more of annual income, education level, family size, or job category of the users.
This invention relates to a data processing system that groups user data sets into clusters based on demographic and socioeconomic factors. The system collects and analyzes user data, including attributes such as annual income, education level, family size, and job category, to categorize users into distinct clusters. These clusters represent groups of users with similar characteristics, enabling targeted analysis or personalized services. The system may also incorporate additional user data, such as geographic location or behavioral patterns, to refine the clustering process. By leveraging these demographic and socioeconomic factors, the system provides insights into user segments, which can be used for marketing, policy-making, or service optimization. The clustering process ensures that users with comparable profiles are grouped together, facilitating more accurate and actionable insights. This approach improves decision-making by identifying patterns and trends within specific user segments, allowing for more effective targeting and resource allocation. The system may be applied in various industries, including finance, healthcare, and education, to enhance personalized services and strategic planning.
13. The system of claim 12 , wherein the grouping of the each of the one or more clusters into a plurality of subclusters is based on one or more of credit score, an account balance, available credit, or percentage of credit used of users in each of the one or more clusters.
This invention relates to a system for analyzing and grouping user financial data to improve decision-making in financial services. The system addresses the challenge of efficiently categorizing users based on their financial behavior and attributes to enable targeted financial products, risk assessment, or personalized services. The system processes financial data from multiple users, including credit scores, account balances, available credit, and the percentage of credit used. This data is used to form one or more clusters of users with similar financial profiles. Each cluster is further divided into subclusters based on additional financial metrics, such as credit score, account balance, available credit, or the percentage of credit used. This hierarchical grouping allows for more granular analysis and segmentation of users, enabling financial institutions to tailor services, assess risk, or detect fraud more effectively. By leveraging these financial attributes, the system provides a structured way to identify patterns and trends within user data, improving the accuracy of financial decisions. The subclustering step ensures that users within each cluster are further refined into smaller, more homogeneous groups, enhancing the precision of financial modeling and risk management. This approach supports better decision-making in areas such as loan approvals, credit limit adjustments, and fraud detection.
14. The system of claim 8 , wherein the plurality of functions further comprises: removing personally identifiable information from each of the plurality of user data sets.
A system for processing user data includes a plurality of functions that analyze and transform data sets. One of these functions involves removing personally identifiable information (PII) from each user data set. The system may also include components for receiving user data from multiple sources, normalizing the data into a standardized format, and applying machine learning models to extract insights. The PII removal function ensures compliance with privacy regulations by anonymizing or de-identifying sensitive information such as names, addresses, or identification numbers before further processing. This function may use techniques like tokenization, hashing, or redaction to obscure PII while preserving the utility of the remaining data. The system may also include validation mechanisms to verify that PII has been effectively removed and that the anonymized data remains useful for analysis. The overall system is designed to handle large-scale data processing while maintaining data privacy and regulatory compliance.
15. A computer-implemented method for training and using a neural network to determine relevant resolutions, the method comprising: receiving or generating, by one or more processors, a plurality of user data sets, wherein each user data set of the plurality of user data sets comprises a user identification data of a user and a detailed user data of the user; removing, by the one or more processors, personally identifiable information from each of the plurality of user data sets; grouping the plurality of the user data sets, by the one or more processors, into one or more clusters of user data sets; grouping, by the one or more processors, each of the one or more clusters into a plurality of subclusters; for each of the plurality of subclusters, training, by the one or more processors, a neural network to associate the subcluster with one or more sequential patterns found within the subcluster, based on the user data sets in the subcluster, to generate a trained neural network; receiving, by the one or more processors, a first series of transactions of a first user; inputting, by the one or more processors, the first series of transactions into the trained neural network; classifying, using the trained neural network, by the one or more processors, the first user into a subcluster among the plurality of subclusters, based on the first series of transactions input into the trained neural network, to generate a classified subcluster of the first user; based on the classified subcluster of the first user, determining, by the one or more processors, whether a model user associated with the first user should be identified; based on the determination of whether the model user associated with the first user should be identified, searching, by the one or more processors, the plurality of user data sets for the user data set of the model user, wherein the user data set of the model user (i) is grouped into one of one or more model subclusters, wherein the one or more model subclusters is predetermined to be model subclusters among the plurality of subclusters, and (ii) comprises a second series of transactions having one or more transactions with identical attributes, compared to attributes of one or more transactions of the first series of transactions; based on both the first series of transactions and the second series of transactions, generating, by the one or more processors, an indicator of a resolution associated with the first series of transactions; converting, by the one or more processors, the indicator into a natural language statement; and displaying, by the one or more processors, the natural language statement on a device associated with the first user.
This invention relates to a computer-implemented method for training and using a neural network to determine relevant resolutions for user transactions. The system addresses the challenge of identifying and resolving transaction-related issues by leveraging user behavior patterns. The method begins by processing user data sets, each containing user identification and detailed transaction data. Personally identifiable information is removed to ensure privacy. The data sets are then grouped into clusters and further divided into subclusters based on transaction patterns. A neural network is trained for each subcluster to recognize sequential transaction patterns within that group. When a new user's transaction series is received, the system inputs it into the trained neural network to classify the user into the most relevant subcluster. If the user's subcluster is identified as a model subcluster (a predefined group of users with similar transaction behaviors), the system searches for a model user whose transaction data closely matches the new user's transactions. By comparing both transaction series, the system generates an indicator of a resolution for the new user's transactions. This indicator is converted into a natural language statement and displayed to the user, providing a clear and actionable solution. The approach enhances transaction resolution by leveraging machine learning to identify and apply relevant patterns from similar users.
Unknown
May 26, 2020
Browse 5M+ US patents with plain-English claim translations and AI-generated analysis.